738
NICHOLAS A. MILASAND
CHARLES R. MILONE
Vol. 68
with an additional 150 cc. of ether. The combined ether solutions were dried with magnesium sulfate. The drying agent and ether were removed, and the residue was distilled from a 50-cc. Claisen flask. The kdimethylaminostyrene boiled a t 85-90' (2.5 mm.). If any dehydration occurred on distillation the water was separated, the product redissolved in ether and dried again. Redistillation in a three-plate, helices-packed column gave a boiling point of 90-91' (2.5-3 mm.); n% 1.6010. The yield was 11 g. or 45T0 of the theoretical amount. Several experiments produced lower yields with a large amount of polymeric residue. A high-boiling polymeric residue could not be distilled. Anal." Calcd. for C&IN: C, 81.58; H, 8.89. Found: C, 81.44; H, 8.91. A picrate was obtained as a yellow powder from 95% ethanol, m. p. 93-94'. If the product from the decomposition of the Grignard complex was not distilled, a small yield of white crystals was obtained, which, when recrystallized from petroleum ether, belted at 58.5-59". Sachs and Sachsl* report 60'
benzoyl peroxide, and 15 cc. of purified dioxane. This was gently refluxed for twenty-four hours. The cooled solution was slowly dropped into 200 cc. of methyl alcohol with vigorous stirring. No powder or gum precipitated, and the sblution was then evaporated t o dryness. A heavy gum resulted, completely soluble in benzene. An attempt to precipitate this gum as a powder failed. The following table summarizes the data on these polyners.
8.37. Polymerization of m-Trifluoromethylstyrene and mMethylstyrene.-In a Pyrex test-tube was placed 1 g. of the monomer. The test-tube was suspended under an ultraviolet lamp and left there until a hard polymer had been formed (twenty-four hours). The polymer was dissolved in 60 cc. of benzene and precipitated by slowly dropping the solution into 300 cc. of methanol with vigorous mechanical stirring. This process was repeated and the powder obtained was dried for one week in a vacuum desiccator. Polymerization of $-N,N-Dimethylaminostyrene. (A) Bulk.-About 1 g. of p-N,N-dimethylaminostyrenewas heated for twenty-four hours a t 120" in a small test-tube. At the end of this period a hard resin, completely soluble in benzene, had formed. (B) Solution.-In a 25-cc. Erlenmeyer flask Was placed 0.8 g. - of -$-N.N-dimethvlaminostvrene. . a small amount of
These approximate molecular weights were determined by viscosity measurements with the use of an equation developed by Kemp and Peters for the determination of the molecular weight of polystyrene using the K value for styrene (Kemp and Peters, I d . Eng. Chcm., 34, 1097 (1942)l.
for P-N,N-dimethylaminophenylmethylcarbinol. Anal. Calcd. for Q&ON: N, 8.39. Found: N,
(14) Microanalyses by Miss Theta Spoor, University of Illinois.
TABLE I Polymer, -styrene
Approx. mol. wt.' Softening point, 'C. Solubility in benzene Empirical formula
m-Tri8uoromethyl 7346 130-155
+
COHIPI
C 62.79 N ... C 62.73 N
m-Methyl-
33000 ii7-12a
+
CiHm 91.46 8.54
...
90.08 8.26
...
...
P.N,N-Dimethylamino- bulk 1620 80-95
+
CIOHIIN 81.58 8.89 9.51 80.45 8.09 8.77
Summary 1. The preparation of m-trifluoromethyl-styrene is described. New procedures for the preparation of m-methylstyrene, m-nitrostyrene and p-N,N-dimethylaminostyrene have been developed. 2. The polymers of these monomers, except m-nitrostyrene, have been prepared and characterized. URBANA, ILLINOIS
RECEIVED JANUARY 15, 1946
[CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, MASSACHUSETTS INSTITUTE OF
TECHNOLOGY ]
Studies in the Synthesis of the Antirachitic Vitamins. 11. A6-Androstenol-3 and its Partial Dehydrogenation with Benzoquinone1a2 B Y NICHOLAS A. MILASAND CHARLES R. MIL ONE^
The simplest sterol which would have the same tdined which contained 16% of A6s7-andronuclear structure as that found in the anti- stadiene-3-01. Irradiation of this crude product rachitic provitamins is A.6~7-androstadiene-3-ol.under standard conditions produced a mixture This could be prepared from As-androstenol-3, which was biologically inactive a t a level of 4000 either by the method of Windaus, Lettrk and U. S. P. XI units of vitamin D per gram6 Ass7Schenck* or that of Milas and Heggie.6 When Androstadiene-3, 17-dioil7 another simple sterol the latter method was used, a product was ob- containing the provitamin D nuclear structure, was found to form no antirachitic products upon (1) Last paper, Milas and Alderson. Jr., THISJOURNAL, 61, 2534 (1939). irradiation.sJ Two explanations may be ad(2) The present paper was originally submitted on March 20, vanced for the failure of these simple substances 1940. to form antirachitically active products upon (3) Abstracted from a portion of a thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy irradiation. Much longer time of irradiation in Chemistry, M. I . T . , August, 1939. Present address: Goodyear than that used under standard conditions may be Rubber Company. (4) Windaus, LettrC and Schenck, Ann., 630, 98 (1935). (5) Milea and Heggii. THIS JOURNAL, 60, 984 (1938); Sah, Res. trov. clim., 69, 454 (1940); cf. Criegee, Ber., 69B, 2758 (1036); Arnold and Collins, THIS JOURNAL, 61, 1407 (1939).
(6) Reported to the authors by Professor Robert S. Harris. (7) Butenandt, Hausemann, Dresler and Meinster, Bcr., 71B, 1316 (1938). (8) Dimroth and Paland, ibid., TPB, 187 (1939).
DEHYDROGENATION OF A6-ANDROSTENOL-3
May, 1946
necessary to cause the opening of ring B. It may also be possible that the side chain plays a specific biological r81e in imparting antirachitic properties to the rest of the molecule, and even if ring B were opened the substance formed without the side chain may be devoid of antirachitic properties. Attempts to prepare A5-androstenol-3through a sequence of reactions by reducing catalytically dehydroandrosterone acdtate, replacing the hydroxyl group on carbon seventeen by bromine and subsequently replacing this by a hydrogen atom were unsuccessful, since we were unable, after several trials, to reduce the keto group by the method of Ruzicka and W e t t ~ t e i n . ~ This was finaSly prepared by way of the WolffKisliner reduction.'O While our work was under way, Raoul and Meuniers" used the same reduction for the same purpose but the product obtained did not agree in properties with our preparation. Since our work, Heard and McKay12aand Butenandt and Suranyi12bhave made the same product by the sa.nie procedure. The physical properties of our product and the acetate derived from i t correspond c l o d y to those given by Heard and hlcKay. TABLE I PHYSICAL
PROPERTIES OF h6-ANDROSTENOL-3 AND DERIVATIVES
Investigator
De Fazzi and Pirrone Raoul and Meunicr Heard and McKay Butenandt and Suranyi Present authors a Analysis reported.
------Melting Alcohol
points, Acetate
202-203" 104" 137 131" 136-138"
185-186" 114 94 91-93' 92-94'
O C . -
Benzoate
.. .. . .....
.
.
I
.
.
. . .. . 173-174'
Experimental Ab-Androstenol-3 (cis 3: 10).-The setnicarbazone (1.15 9.) of dehydroandr~sterone'~ was heated in a sealed tube for seven hours a t 180-185' with a solution of 14 cc. of absolute ethanol containing 1 g. of sodium. The mixture was then poured in 90 cc. of water, neutralized with dilute sulfuric acid and extracted with ether. The gummy material obtained from the ether extract, was dissolved in 10 cc. of absolute alcohol and to the solutioii added 1 g. of Girard reagent P , 1 * 1 cc. of glacial acetic acid, and the mixture refluxed for one hour, then poured in g0 cc. of ice water containing an amount of alkali equivalent to 0.9 of the acidity of the solution. The non-ketonic oily precipitate was extracted with ether, the ether removed and the residue distilled a t 150" (0.0005 mm.) wh$reby 150 mg. of This melting a white solid was obtained; m. p. 99-105 point could not be improved even after several recrystallizations. The difficulty of crystallizing this product was attributed to the presence in the mixture of traces of the epimeric A6-ai~drostenol-3(trans 3: 10). The desired cis isomer was separated from the mixture by means of its insoluble digitonide, the digitonide of the trans isomer being much more soluble in ethanol. The
.
(9) Ruzicka and Wettstein, Hclu. Ckim. A d a , 18, 1204 (1935). (10) Wolff, A n n . , 894, 86 (1912). (11) Raoul and Meuniers, Cornpf.rend., 107, 681 (1938). (12aj Heard and McKay, J . B i d . Chcm., 140, lvi (1941). (12b) Butenaiidt and Suranyi, Ber., 7 6 , 501 (1942). (13) Butenaiidl, Z. physiol. Chem., 187, 00 (1035). (14) Girard and Sandulesco, Hclu. Chim. A d o , 19, 1005 (1930).
739
digitonide of the cis isomer was decomposed with pyridine and the final product ryrystallized from ethyl acetate (needles); m. p. 136-138 . Anal. Calcd. for C i 0 H ~ 0 C, : 83.2; H, 11.0. Found: C,82.0; H, 11.1. The cis isomer was produced in much higher yields when the Wolff-Kishner reduction was carried out a t a temperature of 165-170" for a slightly longer period of nine hours. A product was obtained which, without purification via the digitonide, had a m. p. of 130-1320" and a mixed m. p. with the puriiied material of 133-136 . AS-Androstenol-3-acetate (cis 3: 10) .-This substance was prepared by refluxing AJandrostenol-3 (cis 3: 10) with acetic anhydride. The acetate was recrystallized from methanol; m. p. 92-94". Anal. Calcd. for CZ~H~&?OZ: C, 79.7; H , 10.1. Found: C, 78.9, 79.3; H, 9.7, 9.9. A"-Androstenol-3-benzoate (cis 3 :10).-This ester was prcpared in pyridine from A'-androstenol-3 (cis 3 : 10) and benzoyl chloride. The crude benzoate was recrysta!ized from a 50-50 ether-methanol mixture; m. p. 173-174 . Anal. Calcd. for C.?B&&: C, 82.5; H, 9.0. Found: C, 82.5, 82.4; H, 9.3, 9.3. Dehydrogenation of. A6-Androstenol-3-benzoate with Benzoquinone.6-A mixture of A6-androstenol-3-benzoate (0.5 9.) and benzoquinone (0.28 9.) was heated in a sealed evacuated tube a t 140" for four aiid one-half hours. The reaction product was then dissolved in ether and the ethereal solution shaken first with a 20% solution of sodium hydrosulfide to reduce any unused quinone to hydroquinone, then extracted with a 10% solution of sodium hydroxide to remove the hydroquinone. The final product exhibited in the ultraviolet a strong bluish fluorescence which is characteristic for the presence of a conjugated system similar t o that present in the antirachitic provitamins. The benzoate was then saponified with methyl alcoholic potash and the product obtained purified from methyl alcohol. An ultraviolet absorption spectrum of the final product in absolute ethanol showed the presence of two bands: a prominent band with a maximum at 255 mp having a molecular extinction coefficient of 1650, and a weaker one a t 280-282 mN. A6-Androstenol-3exhibited no selective absorption in this region. A Comparison of the coefficient of the prominent band with the analogous band of ergosterol indicated that A**7-androstadiene-3-olwas present in the mixture t o the extent of about 16%. An attempt to purify the dienol by chromatographic absorption was unsuccessful. Addition of Maleic Anhydride to A6r7-Androstadiene-3benzoate.--A quantity of 0.37 g. of the crude dehydrogenated A6-androstenol-3-benzoate in 4 cc. of dry xylene was heated in a sealed evacuated tube for ten hours a t 135-140' with 0.1 g. of freshly distilled maleic anhydride. The xylene and excess maleic anhydride were then distilled in vacuum aiid the residue saponified with methyl alcoholic potash. After acidification and cstraction with ether, a solid product was obtained which was freed from benzoic acid by a thorough digestion with boiling petroleum ether. The final maleic acid adduct was further purified by precipitating it from chloroform with petroleum ether; m. p. 185-190" (with decomposition). Anal. Calcd. for G3H3205: C, 71.14; H, 8.25. Found: C , 71.7; H, 8.48. Irradiation of A6~7-Androstadiene-3-ol.-Fifty-fivemg. of the crude A~~7-androstadiene-3-ol was dissolved in 190 cc. of anhydrous ethcr and the solution irradiated for thirty minutes in a quartz flask atid in an atmosphere of nitrogen with a quartz-mercury lamp under conditions exactly identical with those used for irradiation of ergosterol. The product was then tested biologically by Professor Robert S. Harris who reported no activity a t a level of 4000 U. S. P. XI vitamin D units per gram.
A'-Androstene-3,17-diacetate.-A6Androstene-3,17-diel (1.85 g.), prepared from dchydroandrosteroiie, was acetylated with acetic aiihydride. Thc crude diacrtate was recrystallized froin nlethanol, yield 1.7 6 . ; 111. p . 158-159'.
740
EMILKAISERAND VIRGILL. KOENIG
which agrees with that reported by Ruzicka and Wettstein.¶ ~~-Androstene-3,17-dibenzoate.-The dibenzoate was prepared by heating on the water-bath a mixture of A~-androstene-3,17-diol, benzoyl chloridc and pyridine. The crude dibenzoate was recrystallized from methanol; m. p. 203-205". Anal. Calcd. for C ~ ~ H ~ BC, O +79.5; : H, 7.6. Found: C, 79.3; H, 7.8. Dehydrogenation of A6-Androstene-3,17-diacetatewith Benzoquinone.-A mixture of 0.3 g. of the diacetatc aiid 0.0!14 g. of berizoquiiione was heated i n a sealed evacuated tulx a t 1:30" for three hours. The product from this rcact i o i l was tlissolvcd i i i ether atid thc solution thoroughly shakeii n i t h a 20yc solution of sodium liydroscilfitc, then estractcd sevcral times with a l0yo sodium hydroxide solutioti, dried, and the ether removed. The slightly tiiscolored rcsidue was purified from nicthanol and a product was obtained the spectrum of which showed a singlc band with a inaxiinilm at 288 nip having a molecular estinctioii coefliciciit of 1116. A6-Aiidrosteiie-3,17-tliacetateesliibitetl no selective absorption in this region. A eompariS O I L of the extinction coelliciciit with that of ergosterol (lO,GOO a t 282 nip) ititlicatctl that the iiiatcrial coiitaitictl the clcsircd dehydrogeriatetl product t o the estciit of about 10%. Irradiation of A5~i-Androstadiene-3,17-diol.-A sainplc OT t Iic iurtially tleliydrogciiiatctl tliacctate was 1iytfrolyzc.ci with iiiciliyl alcohnlic potash arid the iioii-sai,oiiiliable
[CONTRIDUTION FROM T H E
Vol. 68
diol irradiated in ether for four hours using a quartzmercury lamp. The product was theii assayed biologically by Professor R. S. Harris who reported no activity at a level of 10,500 U. S. I?. XI vitamin D units per gram. Almost a year after these experiments were performed, Dimroth and PalandB reported similar biological results with this substance made by a different procedure.
The authors wish to express their appreciation for a generous sample of dehydroandrosterone supplied to one of us (N. A. 11.1.) by the Schering Corporation through the courtesy oi Dr. Erwin Schwenk.
Summary 3 . ~5-Aiitlrosteiiol-3(GiS 3 : 10) aritl its acetate antl benzoa.te were prepared in the pure state. 2. l'he partial dehydrogenation (if h5-antlrostenol-3-benzoate antl &5-androstc:tic-R,li-diacetate with benzoquinone has been studied. The irradiated dehydrogenation products were devoid of antirachitic action when tested at levels of 4000 and 10,500 U. S. I-'. XI vitmiiii D units pcr grain, respcctively. CAMBRIDGE:, MASSACHUSETTS h C E I V E D NOVEMHER
30, 1'3-15
ARMOUR LAUORATORIES]
The Ultraviolet Absorption Spectra of Allyl and Propenyl Substituted Derivatives of Diethylstilbestrol and Hexestrol BY EMILKAISERAND VIRGILL. KOENIC In a previous communication' the preparation of 3,3'-allyl and 3,3'-propenyl diethylstilbestrol and hexestrol are described. The 3,S'-allyl substituted derivatives were rearranged by heating in alkaline solution to coinpounds which were ;I&sui 1 ictl to 1)e :$, '-i)ropcnyl tlcrivativcs in ~tiialog~. to rvactioiis rcpurtctl i i i thc litcriiturc.? Obviously, thc reaction described by Ealbiario3 for thr: cliff crentintiori between allyl and propenyl groups cannot bc applied here inasmuch as the large number of double bonds would introduce complications. For this reason it was decided to exatninc the ultraviolet absorption spectra of these coiiipouiids. An important difference in the structure of the 3,3'-allyl and 3,3'-propenyl tlerivativcs tinder discussinti here resides iii the positioii of tlie tlorible bonds ol' the side chains. In the pr(JT)CllYl roinpountts the double bonds are i n a conjugatccf position to the double bonds of the aromatic iiuclei; i n the allyl derivatives the double bonds are not conjugated. Since the presence of the double bonds is shown by the shape of the ultraviolet absorption curves of organic compounds, i t was thought possible to demonstrate the shifting of the double bonds by comparing the f I ) Kaiser aricl Svarz, 'L'iirs fuunsni.. 68, 036 (1046). ( 2 ) "Organic Keactioiir," Vu!. 11, J o l i i l \Vilcy alltl Suns, Inc., New York, N . Y . , 1941, p. 19. (3) l ~ : l l l , , , m < > l h, , . , 48, 3!bt (101.->).
absorption spectra of 3,3'-allyl derivatives of diethylstilbestrol and hexestrol with the absorption spectra of 3,3'-propenyl derivatives of diethylstilbestrol and hexestrol. Acknowledgment.-The authors are indebted to Professor F. C. Koch for his Iielpftil criticism of the 1ii;iiiuscript and for his eiicouragcmeiit during the work.
Experimental The ultraviolet absorption Curves of the followi~~g coinpounds were determined: diethylstilbestrol, diallyl ether of dicthylstilbestrol, 3,3'-allyl diethylstilbe.stroi, 3 3 ' propenyl diethylstilbestrol and hexestrol. hexestrol diallyl ether, 3,3'-allylhexestrol and 3,:3'-propeiiylhexestrol. The diethylstilbestrol and hexestrol w r r e recrystallized froin commercial preparations. The othcr compouncls were prepared accordiiig to a previous coininunicatioti.' Acetotic-free absolute niethauol, tlistillctl froin alkali, was used as solvent. The coiicciitration of the solutions was 0.00570. The tiieasurenients were made on a Bechman spcctrophotonletcr using the hydrogen discharge tribe as the source of light. Transmittance values were determined at 10 millimicron increments from wave length 220 millimicrons t o 400 millimicrons. Extinction values were obtained from the transmission values and, from these, molecular extinctions were calculated and used for plotting the graphs.
Discussion In Fig. 1 the absorption spectra of diethylstilbestrol, clinllyl ether of tliethylstiibestrol, 3J'-